แบบจำลองอคติความชอบ

You're describing the exact trap Munger warns about: the Liking/Loving Tendency makes you ignore flaws in what you love and distort facts to accommodate it. The fact that you're making excuses internally is the first red flag—your admiration has become a lens, not a mirror.

Run a cold test. Write down the last two launch outcomes in numbers only: revenue, retention, churn, CAC. Then ask yourself: if a stranger presented this same roadmap logic, would you approve it? If the answer is no, your affection is overriding judgment.

Second, separate the person from the decisions. Your co-founder may be brilliant in product vision but terrible at sequencing or market timing. That's not a character flaw—it's a skill gap. But you're treating his calls as extensions of his brilliance, which is the halo effect in action.

Third, get an external voice. Bring in an advisor or board member with no emotional stake. Let them audit the roadmap process, not the person. If they flag the same issues, you have your answer. Loyalty should survive honest feedback—if it can't, it's not loyalty, it's delusion. The question isn't whether you like him; it's whether the business can afford another flop.

You are describing the Liking/Loving Tendency in its purest form: admiration hijacking judgment, then distorting the evidence. The CEO’s taste becomes a halo that wraps the feature, making its flaws invisible to you and your team. Your defense in meetings isn't a design opinion; it's a loyalty reflex.

Run a decoupling test. Before any meeting, write down three concrete, testable reasons this feature wins on its own merits—not aesthetics, not the CEO, not the vision. If you can’t produce them, you’re mirroring, not believing.

Then force the data to speak louder than the admiration. Bring the abandonment curve into the room, not as an attack on the CEO, but as a question: “What would this feature need to change to flatten that curve?” That reframes the discussion from taste to evidence, and it protects the CEO’s ego while exposing the design’s weakness.

Finally, ask yourself privately: if a junior designer you didn’t admire proposed this exact feature, would you defend it? If the answer is no, you’ve found your answer. Separate the person from the product—the halo is yours to remove, not theirs to wear.

You're making a friendship decision dressed in business clothes. The Liking/Loving Tendency is doing exactly what Munger describes: you're ignoring the flaws of what you love and distorting the data to accommodate that love. "Just okay" numbers are your red flag — a stranger with those results wouldn't get a second meeting. Your college history is the halo, and it's bleaching out the performance gap.

Run a clean test: strip the names. Write down the campaign metrics, the audience fit, the conversion cost, and the creative quality — then ask yourself, "If this were a random influencer I met yesterday, would I sign this?" If the answer flips, you have your verdict.

Here's the hard rule: protect the friendship by making it a separate lane. Hire her only if she wins on a blind scorecard. If she doesn't, tell her the truth — "the data didn't clear our bar" — and keep the friendship intact. That's the move a real friend makes, and it's the only move that keeps your P&L honest. Buffett's line applies: don't bet on the jockey when the horse is broken, no matter how much you like the rider.